Devices are now widely available for users to record video and audio of people or events. Often, the user has recorded long segments of video or audio, of which certain portions may be of particular interest. Typically, the user has to review or scroll through the entire video or audio recording to locate the portions of particular interest. This is a time consuming and burdensome process and users with large amount of video or audio recordings face difficulties in processing the recordings to obtain useful portions.